I'm going to talk about the importance of glycemic excursions, focusing on the evidence that postprandial hyperglycemia is a risk factor for macrovascular disease and that its control is essential to achieve hemoglobin A1C goals.

Some frequently asked questions are: Is hemoglobin A1C the most important measure of glycemic control, and what is the impact of glycemic excursions on the development of complications?

Here are my answers. Hemoglobin A1C is the most widely used indicator of glycemic control; it reliably reflects average overall glycemic exposure, but it is a relatively insensitive indicator for cardiovascular risk. I don't believe there's any convincing evidence so far from prospective trials that glycemic excursions play a unique role in the development of complications. What I think is important is the absolute plasma glucose level, whether it be fasting or postprandial. Future studies may show that excursions are important.

What I'm going to review is the evidence that glycemia is a cardiovascular risk factor; I'm going to look at epidemiologic studies, biochemical mechanisms, and intervention trials.

I don't think anyone questions the fact that the most common cause of mortality in people with diabetes is cardiovascular disease. Here are data from death certificates, and in terms of ischemic heart disease and peripheral vascular disease and stroke, you can account for most of the mortality in people with diabetes.

Here we have data from the Multiple Risk Factor Intervention Trial (MRFIT) where they looked at cardiovascular mortality in nondiabetic subjects with various risk factors -- smoking, hypertension, hyperlipidemia, family history -- and for any given number of conventional risk factors, the mortality from cardiovascular disease was markedly increased in people with type 2 diabetes. We have evidence that you've seen already from Dr. Gavin of the United Kingdom Prospective Diabetes Study (UKPDS) study; in that study, people randomized to the intensive treatment group had a reduction in myocardial infarction, which was 0.052. Conventionally, it's got to be less than 0.05; but 0.052 is good enough for me!

Here are follow-up data with the latest hemoglobin A1C levels and there was a significant correlation; for every 1% you reduce the hemoglobin A1C, you can reduce cardiovascular complications by 14%. And there's no threshold; it is a linear, continuous relationship.

Hemoglobin A1C reflects overall glycemic exposure: fasting plasma glucose level and postprandial plasma glucose level. As one's glucose tolerance deteriorates, postprandial glucose levels rise faster than fasting.

These are data from a study we published in the Archives of Internal Medicine where we looked at the relationship between hemoglobin A1C and fasting and 2-hour postprandial levels. This rate of increase is 4 times the rate of fasting, and I think you can appreciate that people who have hemoglobin A1Cs in the normal range, below 6%, have significant postprandial hyperglycemia.

Monnier analyzed the contribution of postprandial and fasting glucose levels to hemoglobin A1C, and what he found was that as hemoglobin A1Cs increased, the proportion due to fasting levels increased. And if we look at people below 7.3%, about 70% of the hemoglobin A1C was due to postprandial hyperglycemia.

This shows data from our own study, and I'll show you a little more of these data later; we had patients in this study spanning the range of 15% hemoglobin A1C down to 4.7%. And as you can see, in people with hemoglobin A1Cs below 6.2%, about 90% of the hemoglobin A1C is due to postprandial glucose levels.

What about postprandial hyperglycemia itself as a risk factor? We have prospective studies about overall glycemia; I showed data from the UKPDS. I didn't have time to make a graphic of the Epidemiology of Diabetes Interventions and Complications (EDIC) study, but you heard from Dr. Gavin that the group randomized to the intensive insulin therapy, when they were followed up, had a 57% reduction in cardiovascular disease. Now that is a tremendous reduction, but what about, specifically, postprandial? Well, we have epidemiologic data and I'm going to show some of those.

First, we have data from the Paris Prospective Study, where they looked at people with different categories of glucose intolerance and their cardiovascular mortality. And I think it's very clear here, people with impaired glucose tolerance, isolated postprandial hyperglycemia, have about a doubling of cardiovascular mortality over a 10-year follow-up.

Here are data from the Funagata study; this is a 7-year follow-up and here we have survival rates. We have people with type 2 diabetes, and we have people with impaired glucose tolerance, which carries almost the same mortality as overt diabetes.

We have the European Prospective Investigation into Cancer in Norfolk, United Kingdom (EPIC-Norfolk) study. This was a simple study of about 3 or 4 years' duration; they simply brought people in, checked their hemoglobin A1C and other conventional risk factors, and then 3 or 4 years later, they checked out who was alive and what others died from. After correcting for other cardiovascular risk factors, people who had hemoglobin A1Cs between 5% and 5.4% had roughly a doubling of risk for cardiovascular death compared with the people with hemoglobin A1Cs less than 5%. Who were these people, 5% to 5.4%? Well, we analyzed people in this range using our database; 67% of people in that range have impaired glucose tolerance.

What about plausible biochemical mechanisms by which hyperglycemia may cause cardiovascular disease?

We know hyperglycemia leads to certain things. You get glycosylation of proteins; you get more glucose going through the polyol-hexosamine pathways. You have generation of free radicals. All of these things will activate protein kinase C and diacylglycerol. And free radicals and glycation will lead to a reduction in nitric oxide, an increase in vascular matrix proteins, an increase in growth factors, an increase in plasminogen activator inhibitor-1 (PAI-1) that will decrease fibrinolysis, and an increase in endothelin, which will raise blood pressure. All of these things lead to increased cell proliferation in vessels, increased coagulability, decreased fibrinolysis, reduced vasodilation, increased permeability, and increased matrix formation. All of these things are involved in atherogenesis.

What about controlled intervention trials? Dr. Gavin and I have already mentioned the UKPDS and the Diabetes Control and Complications Trial (DCCT), studies in type 1 and type 2 diabetes showing that reducing hemoglobin A1C reduces cardiovascular disease. What about postprandial hyperglycemia?

We only have 1 study to date, and this was the Study to Prevent Non-Insulin-Dependent Diabetes Mellitus (STOP NIDDM). It was a study primarily designed to see if an agent like acarbose, which slows the absorption of complex carbohydrates postprandially, could reduce the conversion of people with impaired glucose tolerance to type 2 diabetes. This agent only works on postprandial glucose metabolism; it doesn't alter insulin secretion, doesn't alter insulin sensitivity. Additionally, they looked at cardiovascular endpoints.

Here we have all of the things that they looked at, including a composite: any cardiovascular event; it was reduced by 51% and was statistically significant. This was a study that was only done 3 or 4 years, so we have a prospective randomized intervention trial that shows that controlling postprandial hyperglycemia can reduce cardiovascular events.

Targeting postprandial hyperglycemia is critical if you want to achieve optimal hemoglobin A1C levels. The American Diabetes Association recommends a target of below 7%. The American College of Endocrinology (ACE) and the International Diabetes Federation recommend 6.5%. Both of those are rather generous if we want to wipe out hyperglycemia as a cardiovascular risk. I showed that there's an increased risk in the EPIC-Norfolk study in people who have hemoglobin A1Cs of 5.5%.

You remember the database results I showed you earlier? This is part of that study. In this study, we took 164 people with type 2 diabetes who were in unsatisfactory glycemic control. They had hemoglobin A1Cs above 7.5%. We used a forced titration program to try to get their hemoglobin A1Cs below 7%, and this involved progression of addition of oral agents, bedtime insulin, adding rapidly acting insulin to the bedtime insulin, and ultimately some patients got on a basal-bolus regimen. And we analyzed the people who got below 7% vs the people who didn't get below 7%. Our failures had a hemoglobin A1C of 7.5%; the people who got below 7% had a hemoglobin A1C of 6.5%.

Shown here is the day-long pattern in both those groups; the fasting glucose levels are identical. How do the groups differ? By the postprandial levels. In one group, we failed to get their postprandial excursions reduced, so during the course of the day, there was sort of a stepwise effect. Other people went to bed with a higher blood glucose level. And the difference between these 2 hemoglobin A1Cs could be completely accounted for by their postprandial levels.

We have other evidence that treating postprandial hyperglycemia can reduce hemoglobin A1Cs. Here's a crossover study where patients were either treated with a sulfonylurea or insulin lispro plus a sulfonylurea. Sulfonylurea therapy basically targets the fasting glucose level; you lower the glucose level and values throughout the day will just be lower by how much you lower the fasting. When you add an agent like lispro before meals, you're targeting postprandial hyperglycemia. So we see here that, yes, adding lispro reduced postprandial hyperglycemia; it reduced fasting. Obviously if you go to bed with a lower blood glucose, you're going to wake up with a lower blood glucose, and this resulted in a significant drop in hemoglobin A1C.

What is the role of self-monitoring of blood glucose (SMBG)? Dr. Blonde reviewed the evidence that it is very helpful. If postprandial hyperglycemia is important and you're going to treat it, you have to monitor it. So near-normal hemoglobin A1C levels are strongly influenced by postprandial hyperglycemia, and these levels are associated with substantial risk. SMBG around meals can inform people about day-to-day glycemic variation and allow them to make adjustments in diet and lifestyle. I've had many patients who say, "Wow, I checked myself after that pasta and it really went up high." So they know either they've got to eat less pasta or make an adjustment in their regimen if they're going to do that. And postprandial monitoring can help patients achieve targets, especially when you're using insulin before meals or when you're deciding whether a patient needs specific postprandial treatment, such as one of the meglitinide agents, for example.

Dr. Blonde mentioned one of the studies that he was missing, the Schwedes paper, and these are the data. Subjects were randomized to monitor 6 times a day or not at all, but both groups were given intensive education. And those who self-monitored had a reduction in hemoglobin A1C compared with the other group of about 1%.

How do you make use of SMBG? Dr. Davidson is going to go into details about this, but fasting monitoring would be used to adjust long-acting medications and long-acting insulins before meals, to modify the meal or premeal medications. Postprandials can be monitored to see the effect of different meals and the effect of treatment. Many times, you have patients who have their fasting glucose goals but their hemoglobin A1C is not where it should be. So SMBG, logically, has to be postprandial, but is it just after 1 meal or all meals? Because if it's after 1 meal, you may choose to just add a rapid-acting insulin with that meal and at bedtime; this is particularly important in people who might develop hypoglycemia overnight.

To conclude, hyperglycemia is a continuous risk factor for cardiovascular disease with no apparent threshold. Postprandial hyperglycemia occurs before fasting glucose levels or hemoglobin A1C levels become abnormal. Isolated postprandial hyperglycemia is a cardiovascular risk factor. To reduce the risk for cardiovascular disease, postprandial hyperglycemia should be monitored and treated.